Downloaded from Http:

ëZ± o§êu« fhZ oêu§«hZ±¥ ou±uh± ±§Z±« ¶ou§ «£uhu« «uuh± by RhZ§o GZ§u± F±ñJ B.Sc., Victoria University of Wellington, óþþþ M.Sc. (Hons), Victoria University of Wellington, óþþì A thesis submitted in partial fulllment of the requirements for the degree of Dh±§ P«£í in e Faculty of Graduate Studies (Zoology) e University of British Columbia (Vancouver) Z¶¶«± óþÕó © Richard Gareth FitzJohn, óþÕó Af«±§Zh± Species selection — heritable trait-dependent dierences in rates of speciation or extinction — may be responsible for variation in both taxonomic and trait diversity among clades. While initially controversial, interest in species selection has been revived by the accumulation of evidence of widespread trait-dependent diversication. In my thesis, I developed and applied a number of new likelihood-based methods for investigating species selection by detecting the association between species traits and speciation or extinction rates. ese methods are explicitly phylogenetic and incorporate simple, but commonly used, models of speciation, extinction, and trait evolution; I assume throughout that speciation and extinction can be modelled as a birth-death process where rates depend in some way on one or more traits, and that these traits evolve under a Markov process. In particular, I extended the BiSSE (Binary State Speciation and Extinction) method to allow use with incompletely resolved phylogenies, and developed analogous methods for multi-state discrete traits or combinations of binary traits (MuSSE; Multi-State Speciation and Extinction) and quantitative traits (QuaSSE; Quantitative State Speciation and Extinction). I tested the statistical performance of the methods using simulations, investigating their performance with variation in tree size, degree of resolution, number of traits, and departure from the true model. I used each method to consider a dierent biological question; I found that sexual dimorphism was shortlived but associated with elevated rates of speciation in shorebirds; that solitariness and monogamy are associated with decreased speciation rates in primates (showing that a previous analysis was robust to treating both traits simultaneously); and that body size was a poor predictor of speciation rates in primates. In chapter ¢, I extended this analysis of body size to all mammals, and investigated if within-lineage increases in body size (Cope’s rule) were balanced by species selection against large bodied species. I ii found little support for this hypothesis, with clade-specic dierences in the direction of species selection and idiosyncratic variation in speciation rates. Together, the methods I have developed allow testing of long-standing hypotheses about causes of variation in biological diversity. iii P§uZhu Several chapters from my thesis have been published elsewhere: CZ£±u§ ó has been previously published as: FitzJohn R.G., Maddison W.P., and Otto S.P. óþþÉ. Estimating trait-dependent speciation and extinction rates from incompletely resolved phylogenies. Systematic Biology ¢:¢É¢–äÕÕ. I derived most of the mathematical results, developed the soware, carried out the analysis and wrote the manuscript. Wayne P. Maddison and Sarah P. Otto had the original idea for the “skeletal tree” and “unresolved clades” methods, respectively, and both also helped with writing, editing, and guidance. CZ£±u§ ì has been previously published as: FitzJohn R.G. óþÕþ. Quantitative traits and diversication. Systematic Biology ¢É:äÕÉ–äìì. Sarah P. Otto provided guidance and helped edit the manuscript. CZ£±u§ ¦ has been accepted for publication at Methods in Ecology and Evolution, pending minor revisions. I am the sole author of this manuscript. Sarah P.Otto provided guidance, advice, and helped edit the manuscript. e R package this paper is based on is my own work, but contains methods by Emma E. Goldberg (University of Illinois at Chicago), Karen Magnuson-Ford, and Sarah P. Otto. CZ£±u§ ¢ was coauthored with Nick D. Pyneson (Smithsonian Inistitute) and Sarah P. Otto. I developed the idea, ran the analyses and wrote the rst version of the paper. Nick D. Pyneson provided data and advice about Cetacea. Sarah P. Otto provided guidance, advice, and editing. iv TZfu C±u±« Af«±§Zh± ............................. ii P§uZhu .............................. iv TZfu C±u±« .........................v L«± TZfu« ........................... vii L«± F¶§u«........................... viii Ahëuouu±« ........................x DuohZ± ............................ xiii ÕI ±§o¶h± ...........................× ×.× Detecting species selection...........................£ ×.ö Speciation, extinction, & phylogenies....................å ×.î Structure & contents of this thesis...................... óE «±Z± T§Z±-Du£uou± S£uhZ± b Eì±h± RZ±u« F§ Ih£u±uí Ru«êuo Píuu« ................. ×ÿ ö.× Summary..................................... ×ÿ ö.ö Introduction................................... ×× ö.î BiSSE for complete phylogenies........................ ×® ö.® Likelihood calculations for incompletely resolved phylogenies...... ×å ö.£ Bayesian inference................................ ö® ö.å Numerical methods & application...................... ö£ ö.à Results...................................... öà ö. Application to shorebird data......................... î® ö.Ì Discussion.................................... î ìQ ¶Z±±Z±êu T§Z±« b Dêu§«hZ± ................ ®× î.× Summary..................................... ®× î.ö Introduction................................... ®ö î.î Character evolution & diversication.................... ®® î.® Likelihood calculations............................. ®å î.£ Simulation results................................ £ î.å Application to primate body size data.................... å× v î.à Discussion.................................... åà ¦D êu§«±§uu:C£Z§Z±êu Píuu±h AZí«u« Dêu§«hZ± R................................. àö ®.× Summary..................................... àö ®.ö Introduction................................... àö ®.î e methods................................... àî ®.® e approach................................... à£ ®.£ e MuSSE model................................ à ®.å Simulation test assessing the power of MuSSE............... ö ®.à Closing comments............................... Ìÿ ¢S ¶§êêZ ±u L±±u«±¥ S£uhu« Suuh±,C£u’« R¶u, b MZZ Boí Sñu................................ Ìî £.× Introduction................................... Ìî £.ö Results & discussion.............................. Ì£ £.î Methods..................................... ×ÿî äC h¶« ............................ ××× å.× Some issues with these methods....................... ××î å.ö Future directions................................ ×ö× å.î Conclusion.................................... ×ö® Bf§Z£í ........................... ×ö£ A££uohu« ............................ ×®ö Z S¶££uu±Z§í I§Z± ± CZ£±u§ ó............... ×®ö Z.× Root-state calculations............................. ×®ö Z.ö Character-independent model........................ ×®® f S¶££uu±Z§í I§Z± ± CZ£±u§ ì............... ×®à f.× Single character derivation........................... ×®à f.ö Multivariate character derivation....................... ×®Ì h S¶££uu±Z§í I§Z± ± CZ£±u§ ¦............... ×££ h.× Tuning diversitree................................ ×££ h.ö A faster algorithm for BM & OU likelihood calculations......... ×åö h.î MuSSE & multitrait diversication in primates............... ×åÌ o S¶££uu±Z§í I§Z± ± CZ£±u§ ¢............... ×àÌ o.× Partition compositions............................. ×àÌ vi L«± TZfu« TZfu ì.Õ Summary of model ts for the association between body size and diversication for primates.................. åå TZfu ¦.Õ Summary of model types available in diversitree......... àå TZfu ¢.Õ Properties of the Õþ clades used............... Ìå TZfu ¢.ó Properties of partitions recovered by Muo¶«Z ......... ×ÿ vii L«± F¶§u« F¶§u Õ.Õ Lineage-through-time plots for simulated trees.........à F¶§u ó.Õ Ways that phylogenetic information may be incomplete...... ×î F¶§u ó.ó Schematic of the BiSSE method with and without full phylogenetic knowledge....................... ×£ F¶§u ó.ì Posterior probability densities for BiSSE parameters....... ö F¶§u ó.¦ Posterior probability densities for diversication rates...... îÿ F¶§u ó.¢ Uncertainty around BiSSE parameter estimates as a function of phylogenetic knowledge.................. îö F¶§u ó.ä Uncertainty around diversication rate estimates as a function of phylogenetic knowledge.................. îî F¶§u ó.ß Phylogenetic tree of the ì¢þ species of shorebirds (Charadriiformes). î£ F¶§u ó. Association between sexual and diversication and character transition rates in shorebirds.................... îà F¶§u ì.Õ Possible ways a lineage extant at time t + ∆t might go extinct.... ® F¶§u ì.ó Possible ways a lineage extant at time t + ∆t might lead to exactly the clade N as observed................... £ÿ F¶§u ì.ì Power to detect dierential speciation and extinction with QuaSSE on simulated phylogenies.................. £Ì F¶§u ì.¦ Representative speciation and extinction function ts....... åÿ F¶§u ì.¢ Power to detect trait-dependent speciation and directional character change......................... åö F¶§u ì.ä Phylogenetic tree of the primates............... å® F¶§u ì.ß Primate speciation and extinction model ts.......... å F¶§u ¦.Õ Uncertainty around multitrait-MuSSE parameter estimates as a function of tree size and number of traits............... ® F¶§u ¦.ó Power and error rates of multitrait MuSSE........... £ F¶§u ¦.ì “Main eects” of monogamy and solitariness

Load more